- Chip companies are gaining ground in inference workloads, driven by a shift from GPUs to CPUs in data centers.
- AMD is carving out a role in inference against Nvidia's GPU dominance, while Broadcom gains Apple as a customer for custom AI chips, benefiting Google's growth.
AMD and Nvidia are seeing a shift in AI workloads toward inference, enabled by new chiplet design allowing better memory packaging.
2 reports, 2 independent
Updated Jul 23
Gone quiet
- Reports
- 2
- Developments
- 1
- Repetition
- 50%
New informationRepeats or wire copies
AI-generated analysis. Brind wrote this summary from the reports listed below. It can be wrong. Each section says how much you can rely on it, and the sources are linked so you can check.
What happened
AMD and Nvidia are seeing a shift in AI workloads toward inference, enabled by new chiplet design allowing better memory packaging.
Who's involved
What this event is mainly aboutKeep exploring
Part of
AMD is carving out a role in inference against Nvidia's GPU dominance, while Broadcom gains Apple as a customer for custom AI chips, benefiting Google's growth.Also in this story
- Both AMD and Nvidia provide GPUs for AI inferencing workloads.
- AMD and Ramana Bandili announced a collaboration to advance next generation of AI.
- AMD faces software gaps that limit its ability to rival Nvidia's dominance in the AI chip market.
- AMD and Google are competing in the AI chip and GPU market.
The entities involved
Related events
- Supermicro hosts a summit addressing AI infrastructure bottlenecks and formalizes a new AI memory architecture, CMX, in collaboration with Nvidia.
- AMD, Intel, and Nvidia are engaged in head-to-head competition in high-performance AI workloads, alongside a new partnership between AMD and Rackspace.
- AMD, Nvidia, and SanDisk are involved in AI infrastructure bottlenecks and competition.
- Core Scientific is utilizing GPUs for AI deployments.
- AMD, Intel, Nvidia, and Cerebras are contending in the AI infrastructure market while defending their market share in the CPU space.